E-commerce documentary photography XL version

Bottom mold: Basic algorithm XL

Trigger word: jiadian (optional, tested many images and the difference is not significant)

Approximately 160 carefully selected training materials were used, tested for about three days, the image is relatively stable, simple commercial images can be produced without the need for complex prompt words.

(Images generated through this algorithm can be used for commercial purposes)

Can be combined with other e-commerce lora to enhance specific effects, generally this one is sufficient.

Common parameters are listed below, most of them are default settings~

Sampling method: Euler a

Iteration steps: 30

Lora weight: default 0.8

Enable high-resolution restoration: 1.5~2

Width & Height: 960x1280     1280x960    1024X1360      (Test others by yourself)

Redraw range: 0.3~0.6

Guidance coefficient: 7